Motor Simulation Theories of Musical Beat Perception

Neurocase The Neural Basis of Cognition

ISSN: 1355-4794 (Print) 1465-3656 (Online) Journal homepage: https://www.tandfonline.com/loi/nncs20

Motor simulation theories of musical beat perception

Jessica M. Ross, John R. Iversen & Ramesh Balasubramaniam

To cite this article: Jessica M. Ross, John R. Iversen & Ramesh Balasubramaniam (2016) Motor simulation theories of musical beat perception, Neurocase, 22:6, 558-565, DOI: 10.1080/13554794.2016.1242756 To link to this article: https://doi.org/10.1080/13554794.2016.1242756

Published online: 11 Oct 2016.

Submit your article to this journal

Article views: 235

View Crossmark data

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=nncs20 NEUROCASE, 2016 VOL. 22, NO. 6, 558–565 http://dx.doi.org/10.1080/13554794.2016.1242756

REVIEW Motor simulation theories of musical beat perception

Jessica M. Rossa, John R. Iversenb and Ramesh Balasubramaniam a aCognitive and Information Sciences, University of California, Merced, CA, USA; bSwartz Center for Computational Neuroscience, Institute for Neural Computation, University of California, San Diego, CA, USA

ABSTRACT ARTICLE HISTORY There is growing interest in whether the motor system plays an essential role in rhythm perception. The Received 1 May 2016 motor system is active during the perception of rhythms, but is such motor activity merely a sign of Accepted 22 September unexecuted motor planning, or does it play a causal role in shaping the perception of rhythm? We 2016 present evidence for a causal role of motor planning and simulation, and review theories of internal KEYWORDS simulation for beat-based timing prediction. Brain stimulation studies have the potential to conclusively Auditory perception; motor test if the motor system plays a causal role in beat perception and ground theories to their neural planning; motor simulation; underpinnings. rhythm; timing

Introduction perceptual processes, and this bidirectional causality is a char- acteristic of the models and theories reviewed here. While A tight relationship between movement and auditory rhythm there is a long history of study in how the sensory systems perception is evident in human motor system response and inform action, there is now growing evidence that internal motor involvement during music listening and rhythm tasks forward models make predictions about the sensory conse- (Iversen & Balasubramaniam, 2016; Janata, Tomic, & quences of motor acts (Prinz, 1997; Wolpert & Flanagan, 2009). Haberman, 2012; Repp, 2005a; Repp, 2005b; Ross, These predictions are thought to contribute to sensory per- Warlaumont, Abney, Rigoli, & Balasubramaniam, 2016), and ception and error assessments used for making corrections for can be observed in neural response to music early in infant discrepancies between expected and actual sensory input. An development (Kuhl, Ramirez, Bosseler, Lin, & Imada, 2014). important task for music neuroscience is to understand to How we move to music has by itself become a systematic sub- what extent there is a bidirectional relationship between audi- field of inquiry (Ross et al., 2016) that often focuses on body tory perception and action during not only performance but synchronization with music. Some have suggested that the perception. link between auditory and motor involvement in music could In this review, we focus on the particular relationship be similar to that found in language (Patel, Iversen, & between motor planning and musical beat perception. We Rosenberg, 2006). examine neural and behavioral evidence for active motor Music often impels us to move in time with a perceived involvement in auditory rhythm perception, and contrast this pulse or beat, implying a forward connection between audi- with other general “motor” theories of action, and with motor tory and motor systems that enables sound to guide move- theories of speech perception. In particular, we examine the- ment planning and execution. Auditory training has been ories that posit the role of the motor system as “shadowing” or shown to improve motor performance (Stephan, Heckel, “mirroring” the auditory system and others that suggest that it Song, & Cohen, 2015) and has even been explored for move- may play a more causal or “predictive ” role without which ment rehabilitation in patients with Parkinson’s disease human musical beat perception would be impaired. In addi- (Nombela, Hughes, Owen, & Grahn, 2013; Thaut et al., 1996) tion, we discuss theories that move beyond the more literal and recovery after stroke (Altenmuller & Schlaug, 2013). prediction of sensory consequences of motor acts to a more Interestingly, motor planning regions are active even when abstract role of the motor system in generating temporal merely listening to music with a beat and not moving along. predictions. We propose causal methods, such as using tran- This raises the question: is the motor system necessary for scranial magnetic stimulation (TMS) protocols, as a necessary beat perception, or is such motor activity a consequence of experimental step to further define the causal role of the beat perception, reflecting unexecuted movement? The for- motor system in auditory rhythm perception. mer view, perhaps surprising at first if one considers “the beat” to be a property of the music itself, is consonant with the idea that perception and movement are intimately coupled in a Beat perception continuously interacting bidirectional perception–action rela- Beat perception refers to the detection of a regular pulse tionship (Gibson, 1966). This perception–action relationship underlying a rhythmic input stream. Beats can be stressed, or depends upon sensory perception to inform motor planning, accented, in regular patterns to help structure the pulse in a but suggests that the motor system may influence active

CONTACT Ramesh Balasubramaniam [email protected] © 2016 Informa UK Limited, trading as Taylor & Francis Group NEUROCASE 559 predicable way, and this structure is referred to as meter Finally, beat perception is subject to influence by motor (Lerdahl & Jackendoff, 1983; London, 2004). The auditory behavior. Overt body movement can improve perception of stream may be as simple as a metronome or as complex as a timing (Manning & Schutz, 2013) and influence perceptual highly layered and time-varying musical work, but the human interpretation of ambiguous rhythms (Phillips-Silver & brain seems to almost automatically seek a simple regularity, Trainor, 2005, 2007). Overt and covert motor activities are the beat, or pulse, which can serve to organize our move- associated with changes to perceptual acuity. Recent studies ments (as in dance, or tapping your foot to music), but also have shown improvements in beat-perception and finger- can organize our perception of time (Hannon, Snyder, Eerola, tapping entrainment to music when subjects were instructed & Krumhansl, 2004; Palmer & Krumhansl, 1990). Two types of to search for the pulse by moving their bodies (Su & Pöppel, timing that are involved in rhythm perception are interval- 2012). Further research has shown that demonstrations of based (absolute) timing and beat-based (relative) timing (Dalla accelerating motion lead to faster perceived tempo of musical Bella et al., 2016; Grube, Lee, Griffiths, Barker, & Woodruff, excerpts (Su, 2012; Su & Jonikaitis, 2011). Taken together, 2010; Iversen & Balasubramaniam, 2016). Interval-based timing these results provide evidence that beat and meter perception is common to humans and non-human primates (Merchant & are shaped by motor activity. However, in this review we will Honing, 2014; Zarco, Merchant, Prado, & Mendez, 2009). Beat- focus largely on the less explored role of the motor system in based timing may be uniquely human among primates (e.g., beat perception when no overt movement is involved. Merchant & Honing, 2014), and has other properties that make it of special interest for motor theories of timing (Grube et al., Motor system activation during passive listening 2010; McAuley & Jones, 2003; Patel & Iversen, 2014). A number of findings support the notion that motor activ- A second strand of evidence suggesting the potential for a ity may play an active role in shaping beat perception. In motor role in beat perception comes from neuroimaging, particular, as reviewed in the following, it has been suggested which has repeatedly shown that parts of the motor planning that beat-based timing relies on the establishment and main- system are active during rhythm perception, even in the tenance of an internal predictive model, and there is support absence of overt movement, particularly for rhythms that for this in demonstrations of anticipation in motor synchroni- evoke a strong sense of beat. In particular, beat perception zation, tempo flexibility, the susceptibility of beat perception engages dorsal premotor cortex, supplementary motor area to willful control, and improved perceptual acuity of events (SMA), pre-SMA, basal ganglia, and lateral cerebellum that occur on the beat. (Bengtsson et al., 2009; Chen, Penhune, & Zatorre, 2008; Many empirical studies have concluded that beat percep- Grahn & Brett, 2007; Stupacher, Hove, Novembre, Schütz- tion is anticipatory in nature (Miyake, 1902; Repp, 2005b; Bosbach, & Keller, 2013). Although these activations are dis- Woodrow, 1932). Analyses of finger tapping movements that tributed across multiple regions, they are all in areas of the are synchronized with an auditory rhythm demonstrate that brain associated with motor preparation and support that beat taps often temporally precede the beat, an effect called nega- perception engages both motor and secondary motor tive mean asynchrony; humans spontaneously generate structures. expectations of the timing of rhythmic components What is motor activity doing when listening without mov- (Aschersleben, Gehrke, & Prinz, 2001; Drewing, Hennings, & ing? A parsimonious interpretation of this coactivation of Aschersleben, 2001). Another property of beat-based timing is motor regions while listening to rhythm is that it is related flexibility in tempo. Rhythms can speed up or slow down and, to anticipatory movement preparation. Consistent with this despite these temporal fluctuations, people perceive an “mere” motor-preparation view, corticospinal excitability is underlying rhythmic structure. In contrast to synchronous modulated by listening to music with a strong beat. sound production in other species, which is not as demon- Stupacher et al. (2013) measured motor excitability during strably flexible, humans can entrain movements to a range of passive listening using TMS and demonstrated that excitability tempi between 94 and 176 BPM (Hanson, Case, Buck, & Buck, was time locked to the beat, and the degree of excitability 1971; London, 2004; McAuley, Jones, Holub, Johnston, & Miller, reflected auditory–motor training, being greater in trained 2006; Patel & Iversen, 2014; van Noorden & Moelants, 1999). musicians. Motor excitability in amateur pianists while listen- An important feature of musical beat is that it is a percep- ing to a piano piece increases after learning to play that piece tual construct, influenced by but not uniquely determined by of music on the piano (D’Ausilio, Altenmüller, Belardinelli, & rhythms. It is susceptible to conscious control and active Lotze, 2006), and motor excitability is sensitive to differences metrical interpretation on the part of the listener (Iversen, in rhythmic properties between musical excerpts (Stupacher Repp, & Patel, 2009); the sense of beat actively shapes the et al., 2013). perception of rhythm. In the aforementioned study (Iversen However, accumulating evidence suggests motor system et al., 2009), when asked to impose different metrical inter- activation while listening to rhythms may not merely be an pretations onto a rhythmically ambiguous phrase, subjects’ epiphenomenon of suppressed movement, but may also play magnetoencephalography (MEG) recordings reflected ima- a causal role in shaping rhythm perception. One clue comes gined metrical structures despite physical stimulus invariance. from cases where dysfunction of motor regions impairs Metrical interpretation influences early evoked neural rhythm perception. Patients with impaired basal ganglia func- responses in the beta range, with a stronger response on the tion due to Parkinson’s disease (PD) show impairments in a imagined beat, and these patterns resemble those of non- rhythm discrimination task compared with age-similar control imagined physical accents (Iversen et al., 2009). subjects (Grahn & Brett, 2009). In the abovementioned study, 560 J. M. ROSS ET AL. both PD and healthy participants were presented with a dis- pitch even if they show no impairment in pitch discrimination crimination task using beat-based rhythms or non-beat-based tasks (Dalla Bella, Giguère, & Peretz, 2007). Pfordresher’s(2011) rhythms. While there was no difference in discrimination abil- explanation for poor-pitch singing is a deficit in creating an ity between the groups for non-beat-based rhythms, discrimi- inverse internal model from perception of a pitch that can then nation of beat-based rhythms was reduced in the PD group be used for pitch production. This vocal imitation weakness suggesting that the basal ganglia are important for perception has also been demonstrated in intonational speech of musical beat, perhaps for generating an internal beat struc- (Pfordresher & Mantell, 2009). Accurate predictions about ture (Grahn & Rowe, 2013). Although further research is internal and external sensory effects of action are needed for needed to determine how direct and causal the basal ganglia’s skilled movement (Wolpert & Flanagan, 2009), including those involvement is in beat perception, recent work of Kotz, Brown, underlying the articulatory processes in singing. In addition, and Schwartze (2016) suggests that the basal ganglia’s role in predictions cannot be static but instead allow for online beat perception might stem from some aspect of motor pre- updating; errors between predictions and sensory conse- paration or planning. quences are continuously translated into changes in the inter- Additional support for motor planning involvement in audi- nal model (Wolpert & Flanagan, 2015, 2009; Wolpert & Kawato, tory perception can be found in MEG and electroencephalo- 1998; Yang, Wolpert, & Lengyel, 2016). Mere shadowing does graphy studies of beta band neural oscillations. Beta band not support this informative and flexible interchange between modulation is thought to be related to anticipatory proces- action and perceptual consequences via error-based modifica- sing, as beta activity decreases just after tone onset but its tion to the internal model. There are clear advantages to rebound may reflect sequence tempo (Fujioka, Trainor, Large, having a bidirectional predictive basis to connect the auditory & Ross, 2012). The same study found coactivation of auditory and motor systems. and motor cortical areas, even without a motor response, Other domain-general frameworks have been developed, although causality could not be addressed. Other results sug- including ideomotor theories (Shin, Proctor, & Capaldi, 2010) gest that auditory beta-band modulation is influenced by top– and common-coding. Common-coding approaches present down processes: voluntary metrical interpretation of rhythms perception and action as having common representation modulates beta-band responses to sound (Iversen et al., 2009). (Prinz, 1997), thus making claims about the predictive nature Such modulation is suggested to reflect ongoing motor plan- of the relationship between the auditory and motor system. ning processes. According to the common-coding accounts, actions are coded as the perceived effects of those actions (action effects). Thus viewing another person moving activates these action- Motoric basis of sensory prediction perception representations, allowing for perceptual prediction A third field of discourse that relates to the role of the motor generation. The theory is supported by evidence showing system in beat perception is extensive work on internal motor shared neural substrate for perceived and actualized move- models. From the perspective of this work, putative internal ments, and by interference when the two try to access this models used for auditory expectation could be understood as representation simultaneously (Prinz, 1997), reminiscent of either forward or inverse. Forward internal models, such as Gibson’s( 1966) account of perception being in service of efference copies, are used to predict sensory outcomes result- generating opportunities for action and vice versa. ing from motor behavior. Inverse internal models are used to In a recent article, Press and Cook (2015) argued that plan motor behavior based on desired sensory outcomes domain-general motor contributions to perception undermine (Miall, 2003; Pfordresher, 2011; Tian & Poeppel, 2010). We the theory that motor activation while watching human move- can illustrate these types of models and their roles in coordi- ment is for action simulation: that it is only shadowing. They nating action and perception with studies of singing. Studies describe a number of domain-general motor contributions to of internal models involved in singing have used the impact of perception, including recognition of simple movement para- altering perceptual feedback to perturb ongoing production, meters (direction, position, velocity), timing, inference about presumably through the mismatch with an existing forward human motion in masked point-light displays, mental rotation, internal model and a disrupted inverse internal model crea- and visual search. The authors classify these as domain- tion. This has been shown with internal models for vocal pitch general because they are not necessarily related to complex (Pfordresher, 2011) and volume production, known as the motor actions such as grasping; these contributions could be Lombard effect (Lombard, 1911; Zollinger & Brumm, 2011), understood as generically relating to movement. However, and fingertip force production (Therrien, Lyons, & this evidence does not negate that these contributions are Balasubramaniam, 2012). Disruption of the process involving used to form sensory predictions. inverse internal models is thought to be the reason some Although there is considerable evidence for (Rizzolatti & people consistently sing pitches that are too high or low Craighero, 2004) and controversies surrounding mirror neuron when trying to match pitch (Pfordresher, 2011). This phenom- theories (Hickok, 2009), it has been argued that mirror neurons enon is referred to as poor-pitch singing, and has been shown might play an important role in generating inverse and forward to not be attributable to perceptual deficits in pitch percep- internal models (Miall, 2003). Mirror neurons are best known tion, motor deficits, or pitch memory deficits (arguably, as for their activity during visual observation, but can also outlined in Hutchins & Peretz, 2012). The majority of the become active when hearing an action without seeing the general population can carry a tune with pitch and timing action (Kohler et al., 2002). Although there are obvious paral- proficiency, but a small percentage cannot reliably match lels between the theories of motor simulation and mirror NEUROCASE 561 neurons (Koelsch, 2012), the mirror neurons associated with communication between auditory regions and motor planning the ventral premotor area do not appear to be related directly regions. This is related to the concept of reentry, “aprocessof to more dorsal premotor areas that are associated with beat temporally ongoing parallel signaling between separate maps perception. along ordered anatomical connections” (Edelman, 1989). Contrary to the mere shadowing accounts, the idea that According to ASAP, (1) neural signals from auditory to motor the motor system may influence auditory cognition has been planning regions provide information about the timing of audi- present in the literature for some time (Bolton, 1894), and has tory events; (2) these signals influence the timing of periodic recently been discussed by a number of authors (Arnal, 2012; motor planning signals in motor regions, and (3) these planning Jeannerod, 2001; Prinz, 1997; Rauschecker, 2011; Repp, 2005b; signals flow from motor regions back to auditory regions to Schubotz, Friederici, & von Cramon, 2000; Sperry, 1952; Vuust, provide a signal that predicts upcoming beat times. In forward Ostergaard, Pallesen, Bailey, & Roepstorff, 2009; Zatorre, Chen, models such as predictive coding, primary information processing & Penhune, 2007). Recent accounts of sensory gain during operates on predictions of sensory consequences, but in simula- movement support top–down motor influences on sensory tion-based models such as ASAP, top–down (anticipatory) and state (Niell & Stryker, 2010; Nozaradan, Schönwiesner, Caron- bottom–up (reactive) processes work in parallel, continuously Desrochers, & Lehmann, in press; Wekselblatt & Niell, 2015). influencing each other. Andy Clark, Karl Friston and colleagues propose a predic- Other extant models of beat perception that posit top– tive coding model that describes motor behavior as a way of down influences on auditory processing include dynamic selecting sensory input (Clark, 2015). In this framework, infor- attending theory where attention is modulated with temporal mation flow is driven by top–down sensory predictions about event structure (Jones, 1976), and hypotheses relating motor proprioception and other sensory effects, and the only bottom influence on auditory processing to active suppression during up information is in the form of prediction errors. This active vocalization (Arnal, 2012). Nonlinear oscillator models suggest inference, or action-oriented predictive processing, proposes one way this might be achieved by entrainment of hypothe- that downward connections from motor cortex (as from sen- sized neural oscillations with rhythmic auditory events with sory cortex) carry predictions of sensory effects that are only reciprocal interactions among several layers of the network met with bottom up prediction errors. This framework puts required to predict the beat (Large, Herrera, & Velasco, 2015; forth a forward model that is corrected when confronted by Large & Jones, 1999; Large & Snyder, 2009). unexpected sensory consequences. It bypasses the need for inverse models and efference copies in favor of error modu- The path ahead lated corollary discharge (encoded sensory predictions), and is a low cost strategy with minimal computational demands The evidence reviewed above can be organized into two (Clark, 2015). perspectives: motor system activation while listening to Another theory, the “ASAP” (Action Simulation for Auditory rhythms is (1) only shadowing or (2) it also has a predictive, Prediction) hypothesis of Patel and Iversen (2014)makesa causal role in beat perception. Much of the evidence is strong claim for a necessary predictive role of the motor system: suggestive of a causal role, but many questions remain activity in the motor planning system is necessary for beat- that need to be answered to move forward with this work. based perception, and fundamentally shapes our perception of How might we make further progress on these questions? events via connections in the dorsal auditory pathway enabling Below we remark on other motor theories of perception and premotor, parietal and temporal cortices to interact. The ASAP how they do and do not contribute to advancing support of hypothesis suggests that the motor planning system uses the either of these perspectives. We then make experimental same neural machinery involved in simulation of body move- suggestions for causal studies needed to test the validity ment (e.g., periodic movement patterns) to generate or entrain of these perspectives. Further work is required to clearly its neural activity patterns to the beat period, and that these elucidate the role of the motor planning regions in the patterns are communicated from motor planning regions to auditory dorsal stream that could also help distinguish auditory regions where they serve as a predictive signal for the between how the brain responds differentially to music timing of upcoming beats and shape the perceptual interpreta- and other acoustic stimulation like speech (Hickok, tion of rhythms. This hypothesis expands on an earlier sugges- Buchsbaum, Humphries, & Muftuler, 2003). tion by Iversen et al. (2009) that in beat perception the motor system affects the auditory system by injecting precisely-timed Motor simulation theories beat related modulations, which itself was based on earlier psychological suggestion that the beat may involve “covert There is a long history of discussing motor involvement in action” (Repp, 2005b). In contrast to “mirroring” theories, and speech perception, and we might turn to it for comparison to the motor theory of speech perception (discussed below), and contrast. Much as there is motor activation when listening under ASAP the putative motor planning timing signals may, to rhythms, numerous studies show neuroimaging evidence of but need not, be related to imagery of movements of the type motor activation while participants listen to speech (Skipper, that would be required to create the perceptual input. Instead, Nusbaum, & Small, 2005; Wilson, Saygin, Sereno, & Iacoboni, they may be purely abstract timing, possibly, but not necessarily 2004). In addition, there is a range of evidence to suggest that coupled to a specific action. speech effector muscles show facilitation when listening to The central neuroscientific claim of the ASAP hypothesis is that speech (Fadiga, Craighero, Buccino, & Rizzolatti, 2002). MEG beat perception involves temporally precise two-way analysis of infants at ages 7–11 months supports that motor 562 J. M. ROSS ET AL. activation while listening to speech is present in infants that in the dorsal auditory stream. In addition, the temporal reg- are just learning to make pre-speech sounds (Kuhl et al., 2014). ularity of rhythmic contexts could enable prediction in a way It would appear that this multisensory and multimodal rela- that naturalistic speech might not. Beats have a more predict- tionship is present very early in speech development, and that able structure in a way that speech, with all its irregularities, it is likely an integral part of speech perception. does not. For this reason, beat perception might allow for Some accounts of speech perception, such as the Motor more motor simulation than speech perception, and beat Theory of Speech Perception (Galantucci, Fowler, & Turvey, perception paradigms might be more ideal for investigations 2006; Liberman & Mattingly, 1985) and Analysis by Synthesis of perceptually relevant motor recruitment. (Stevens & Halle, 1967), propose that speech perception relies on motor estimation or expectation. The Motor Theory of Causal studies of neural circuits during beat Speech Perception was proposed to address the problem of perception perceptual invariance (Liberman & Mattingly, 1985), which makes its motivation quite distinct from motor simulation TMS is a technique that uses magnetic field pulses applied to theories of rhythm perception. Invariance is the observation the surface of the scalp to cause functional changes in the that a speech signal can have considerable acoustic variation electrical neural activity in superficial cortical regions (Huang, due to context, the speaker’s gender and vocal qualities, and Edwards, Rounis, Bhatia, & Rothwell, 2005). TMS protocols use noise but listeners are still able to group speech sounds into parameters such as magnetic field power and pulse frequency meaningful categories such as phonemes. Liberman and col- to induce temporary excitement or disruption to target corti- leagues proposed that speech sound categories are derived by cal regions, and therefore can be used to change cortical inferring the neural representation of the gestures that pro- activity during or before asking participants to do behavioral duced the sound (Liberman & Mattingly, 1985). experiments to see if changing cortical activation leads to The Motor Theory of Speech Perception has been criticized changes in behavior (Huang et al., 2005). for a number of reasons including its relation to theories of TMS protocols have been used to explore motor theories of speech modularity and lack of specification about how acous- speech. TMS-induced disruption of premotor cortex has been tic signals are mapped to gestures (Sussman, 1989). Another shown to disrupt speech perception (Meister, Wilson, Deblieck, criticism of the Motor Theory of Speech Perception and motor Wu, & Iacoboni, 2007), but Stasenko and colleagues argue that theories in general is that patients with damage to the motor the spread of TMS-induced changes might lead to disruption system can exhibit normal action recognition (Stasenko, of sensory regions in addition to premotor targets. Figure Garcea, & Mahon, 2013), and normal phonemic discrimination, eight TMS coil designs advertise focal stimulation, but the although explicit labeling of speech sounds is impaired spread of activation is not well understood. However, even (Stasenko et al., 2015). Stasenko et al., posit a more nuanced in studies of clinical lesions we see support of motor theories. view that motor representations may be called upon when There is some existing work with TMS showing, using other language cues are not present (such as semantics or causal designs, the neural substrate involved with timing abil- context). ities. Low-frequency repetitive TMS applied over left dorsolat- Although there are similarities between motor theories for eral premotor cortex (dPMC) can interfere with accuracy on a speech perception and motor theories for beat perception, finger tapping synchronization task to an auditory metro- there are also many clear differences. Perhaps foremost, in nome, and this disruption in accuracy occurs whether the motor theories of speech, sounds are mapped to motor repre- participant is tapping with their right or left hand (Pollok, sentations the perceiver would use to produce the same Rothkegel, Schnitzler, Paulus, & Lang, 2008). Because synchro- sounds. In complex music, at least for non-musicians, such a nization employs beat-based predictive timing mechanisms, it direct mirroring is inconceivable. Speech perception relies on may be concluded that left dPMC is involved in beat-based linguistic context in a way that beat perception does timing. Continuous theta burst stimulation (cTBS), a TMS pro- not. Second, as mentioned above, the motivations behind tocol that down-regulates cortical activity at the focal target the theories are distinct. Motor activity in speech perception location (Huang et al., 2005), interferes with interval-based was proposed as a mechanism for creating speaker invariance timing when applied over medial cerebellum, but does not of speech perception, whereas motor involvement in rhythm interfere with beat-based timing (Grube et al., 2010). This perception is proposed as a source of temporally precise supports a functional dissociation between interval and beat- signals to modulate rhythmic expectation and grouping, as based timing, and suggests that cerebellum is involved in well as implement the observed willful endogenous influences interval, but not beat-based, timing. on rhythm perception. Third, the neural circuits implicated in Although there is a scarcity of causal work on neural con- motor theories of speech are distinct from those proposed by tributors to beat-based timing, the weight of these studies is motor simulation theories of beat perception. Speech listening considerable due to strengths of the causal designs. Additional has been shown to be accompanied by bilateral activations in causal work is needed to explore the current motor theories of superior ventral premotor cortex, which are associated with beat perception, and to ground these theories to their neural speech motor production, and in primary motor cortex (Wilson underpinnings. TMS protocols provide powerful causal metho- et al., 2004). Although there is activation during passive dology that can temporarily alter cortical activity in focal motor speech listening in motor areas (Wilson, Molnar-Szakacs, & and premotor regions, either by exciting or suppressing activity Iacoboni, 2008; Wilson et al., 2004), this does not provide (Huang et al., 2005), and this can be used to test theories that support for bidirectional predictive auditory–motor processes claim that the role of the motor system is obligatory for beat- NEUROCASE 563 based timing. For example, disruption of the internal model/ ORCID simulation mechanisms for beat perception should lead to Ramesh Balasubramaniam http://orcid.org/0000-0001-7575-3080 declines in accurate rhythm perception and auditory–motor synchronization. Thus, manipulating motor planning activity and internal model generation in a rhythm task should lead to changes in accurate rhythm perception and production. References Altenmuller, E., & Schlaug, G. (2013). Neurologic music therapy: The ben- eficial effects of music making on neurorehabilitation. Acoustical Conclusions Science and Technology, 34,5–12. doi:10.1250/ast.34.5 Arnal, L. H. (2012). Predicting “when” using the motor system’s beta-band The theories related to speech perception suggest that oscillations. Frontiers in Human Neuroscience, 6, 225. doi:10.3389/ motor activation during speech listening could be mirror- fnhum.2012.00225 ing/shadowing used for auditory processing. The long his- Aschersleben, G., Gehrke, J., & Prinz, W. (2001). Tapping with peripheral nerve block. A role for tactile feedback in the timing of movements. tory of motor theories of speech provide beat perception Experimental Brain Research, 136, 331–339. doi:10.1007/s002210000562 theories with frameworks in which to operate, but do not Bengtsson, S. L., Ullén, F., Ehrsson, H. H., Hashimoto, T., Kito, T., Naito, E. … meet the standard of proof for theories that propose that Sadato, N. (2009). Listening to rhythms activates motor and premotor motor activation during music listening reflects predictive cortices. Cortex, 45,62–71. doi:10.1016/j.cortex.2008.07.002 – processes. Beat-perception and entrainment paradigms, Bolton, T. (1894). Rhythm. The American Journal of Psychology, 6, 145 238. doi:10.2307/1410948 internal model frameworks, and demonstrations of func- Chen, J. L., Penhune, V. B., & Zatorre, R. J. (2008). Listening to musical tional connectivity are contributing to mounting evidence rhythms recruits motor regions of the brain. Cerebral Cortex, 18, for predictive simulations in motor networks. However, 2844–2854. doi:10.1093/cercor/bhn042 these methodologies can only provide suggestive evidence. Clark, A. (2015). Embodied prediction. In T. Metzinger & J. M. Windt (Eds.), Extending these paradigms using causal methodology is Open MIND (Vol. 7). Frankfurt am Main: MIND Group. doi:10.15502/ 9783958570115 needed in order to conclusively show that motor networks D’Ausilio, A., Altenmüller, E., Belardinelli, M. O., & Lotze, M. (2006). Cross- not only shadow speech and music but also provide pre- modal plasticity of the motor cortex while listening to a rehearsed dictive models that can be actively updated and main- musical piece. European Journal of Neuroscience, 24, 995–958. tained. Direct tests of a causal role of the motor system in doi:10.1111/j.1460-9568.2006.04960.x beat perception are needed that perturb the motor system, Dalla Bella, S., Farrugia, N., Benoit, C.-E., Begel, V., Verga, L., Harding, E., & Kotz, S. A. (2016). BAASTA: Battery for the assessment of auditory either through dual task paradigms or through direct neu- sensorimotor and timing abilities. Behavior Research Methods. rostimulation. Motor theories of perception, such as predic- doi:10.3758/s13428-016-0773-6 tive coding and the ASAP hypothesis, propose that motor Dalla Bella, S., Giguère, J.-F., & Peretz, I. (2007). Singing proficiency in the simulation is likely integral to auditory beat-based timing. general population. The Journal of the Acoustical Society of America, 121, – Beat perception studies have beenusedtodemonstratethe 1182 1189. doi:10.1121/1.2427111 Drewing, K., Hennings, M., & Aschersleben, G. (2001). The contribution of strong relationship between motor activation and listening tactile reafference to temporal regularity during bimanual finger to repetitive sounds, and this relationship has possible tapping. Psychological Research, 66,60–70. doi:10.1007/s004260100074 implications for understanding evolutionary origins of Edelman, C. M. (1989). The remembered present: A biological theory of music and its relation to language. Tests of the theories of consciousness. New York, NY: Basic Books. motor involvement for speech perception provide support Fadiga, L., Craighero, L., Buccino, G., & Rizzolatti, G. (2002). Speech listening specifically modulates the excitability of tongue muscles: A TMS for robustness of the link between motor system activation study. European Journal of Neuroscience, 15, 399–402. doi:10.1046/ and auditory processing, and with evidence from beat- j.0953-816x.2001.01874.x perception and internal model paradigms, provide sugges- Fujioka, T., Trainor, L. J., Large, E. W., & Ross, B. (2012). Internalized timing tive evidence for predictive simulation. Further research of isochronous sounds is represented in neuromagnetic beta – should implement causal studies to directly test for predic- oscillations. Journal of Neuroscience, 32, 1791 1802. doi:10.1523/ JNEUROSCI.4107-11.2012 tive motor models in beat perception. TMS is a technique Galantucci, B., Fowler, C. A., & Turvey, M. T. (2006). The motor theory of that can be used non-invasively to investigate the causal speech perception reviewed. Psychonomic Bulletin & Review, 13, relationship between motor simulation and beat-based tim- 361–377. doi:10.3758/BF03193857 ing, and more generally help explore beat perception as a Gibson, J. J. (1966). The senses considered as perceptual systems. Boston, specific example of an action-based perceptual system, pro- MA: Houghton Mifflin. Grahn, J. A., & Brett, M. (2007). Rhythm and beat perception in motor areas viding an example of how the study of music and brain can of the brain. Journal of Cognitive Neuroscience, 19, 893–906. address general mechanisms of brain function. doi:10.1162/jocn.2007.19.5.893 Grahn, J. A., & Brett, M. (2009). Impairment of beat-based rhythm discrimination in Parkinson’s disease. Cortex, 45,54– 61. doi:10.1016/j. cortex.2008.01.005 Disclosure statement Grahn, J. A., & Rowe, J. B. (2013). Finding and feeling the musical beat: No potential conflict of interest was reported by the authors. Striatal dissociations between detection and prediction of regularity. Cerebral Cortex, 23, 913–921. doi:10.1093/cercor/bhs083 Grube, M., Lee, K.-H., Griffiths, T. D., Barker, A. T., & Woodruff, P. W. (2010). Funding Transcranial magnetic theta-burst stimulation of the human cerebellum distinguishes absolute, duration-based from relative, beat-based per- This work was supported by the Division of Behavioral and Cognitive ception of subsecond time intervals. Frontiers in Psychology, 1, 171. Sciences [BCS 1460633]. doi:10.3389/fpsyg.2010.00171 564 J. M. ROSS ET AL.

Hannon, E. E., Snyder, J. S., Eerola, T., & Krumhansl, C. L. (2004). The role of Psychology: Human Perception and Performance, 29, 1102–1125. melodic and temporal cues in perceiving musical meter. Journal of doi:10.1037/0096-1523.29.6.1102 Experimental Psychology: Human Perception and Performance, 30, McAuley, J. D., Jones, M. R., Holub, S., Johnston, H. M., & Miller, N. S. (2006). 956–974. doi:10.1037/0096-1523.30.5.956 The time of our lives: Life span development of timing and event Hanson, F. E., Case, J. F., Buck, E., & Buck, J. (1971). Synchrony and flash tracking. Journal of Experimental Psychology General, 135, 348–367. entrainment in a New Guinea firefly. Science, 174, 162–164. doi:10.1126/ doi:10.1037/0096-3445.135.3.348 science.174.4005.161 Meister, I. G., Wilson, S. M., Deblieck, C., Wu, A. D., & Iacoboni, M. (2007). Hickok, G. (2009). Eight problems for the mirror neuron theory of action The essential role of premotor cortex in speech perception. Current understanding in monkeys and humans. Journal of Cognitive Biology, 17, 1692–1696. doi:10.1016/j.cub.2007.08.064 Neuroscience, 21, 1229–1243. doi:10.1162/jocn.2009.21189 Merchant, H., & Honing, H. (2014 ). Are non-human primates capable of Hickok, G., Buchsbaum, B., Humphries, C., & Muftuler, T. (2003). Auditory- rhythmic entrainment? Evidence for the gradual audiomotor evolution motor interaction revealed by fMRI: Speech, music, and working mem- hypothesis. Frontiers in Psychology, 7, 274. doi:10.3389/fnins.2013.00274 ory in area Spt. Journal of Cognitive Neuroscience, 15, 673–682. Miall, R. C. (2003). Connecting mirror neurons and forward models. doi:10.1162/089892903322307393 NeuroReport, 14, 2135–2137. doi:10.1097/00001756-200312020-00001 Huang, Y., Edwards, M. J., Rounis, E., Bhatia, K. P., & Rothwell, J. C. (2005). Miyake, I. (1902). Researches on rhythmic activity. Studies From the Yale Theta burst stimulation of the human motor cortex. Neuron, 45, Psychological Laboratory, 10,1–48. 201–206. doi:10.1016/j.neuron.2004.12.033 Niell, C. M., & Stryker, M. P. (2010). Modulation of visual responses by Hutchins, S., & Peretz, I. (2012). A frog in your throat or in your ear? behavioral state in mouse visual cortex. Neuron, 65, 472–479. Searching for the causes of poor singing. Journal of Experimental doi:10.1016/j.neuron.2010.01.033 Psychology: General, 141,76–97. doi:10.1037/a0025064 Nombela, C., Hughes, L. E., Owen, A. M., & Grahn, J. A. (2013). Into the groove: Iversen, J. R., & Balasubramaniam, R. (2016). Synchronization and temporal Can rhythm influence Parkinson’s disease? Neuroscience & Biobehavioral processing. Current Opinion in Behavioral Sciences, 8, 175–180. Reviews, 37,2564–2570. doi:10.1016/j.neubiorev.2013.08.003 doi:10.1016/j.cobeha.2016.02.027 Nozaradan, S., Schönwiesner, M., Caron-Desrochers, L., & Lehmann, A. (in Iversen, J. R., Repp, B. H., & Patel, A. D. (2009). Top-down control of rhythm press). Enhanced brainstem and cortical encoding of sound during perception modulates early auditory responses. Annals of the New York synchronized movement. NeuroImage. doi:10.1016/j. Academy of Sciences, 1169,58–73. doi:10.1111/j.1749-6632.2009.04579.x neuroimage.2016.07.015 Janata, P., Tomic, S. T., & Haberman, J. M. (2012). Sensorimotor coupling in Palmer, C., & Krumhansl, C. L. (1990). Mental representations for musical music and the psychology of the groove. Journal of Experimental meter. Journal of Experimental Psychology: Human Perception and Psychology: General, 141,54–75. doi:10.1037/a0024208 Performance, 16, 728–741. doi. 10.1037/0096-1523.16.4.728 Jeannerod, M. (2001). Neural simulation of action: A unifying mechanism Patel, A. D., & Iversen, J. R. (2014). The evolutionary neuroscience of for motor cognition. NeuroImage, 14, S103–109. doi:10.1006/ musical beat perception: The Action Simulation for Auditory nimg.2001.0832 Prediction (ASAP) hypothesis. Frontiers in Human Neuroscience, 8, 57. Jones, M. R. (1976). Time, our lost dimension: Toward a new theory of doi:10.3389/fnsys.2014.00057 perception, attention, and memory. Psychological Review, 83, 323–355. Patel, A. D., Iversen, J. R., & Rosenberg, J. C. (2006). Comparing the rhythm doi:10.1037/0033-295X.83.5.323 and melody of speech and music: The case of British English and Koelsch, S. (2012). Brain and music. West Sussex: Wiley-Blackwell. French. The Journal of the Acoustical Society of America, 119, Kohler, E., Keysers, C., Umilt`a, M. A., Fogassi, L., Gallese, V., & Rizzolatti, G. 3034–3047. doi:10.1121/1.2179657 (2002). Hearing sounds, understanding actions: Action representation Pfordresher, P. (2011). Poor-pitch singing as an inverse model deficit: in mirror neurons. Science, 297, 846–848. doi:10.1126/science.1070311 Imitation and estimation. International Symposium on Performance Kotz, S. A., Brown, R. M., & Schwartze, M. (2016). Cortico-striatal circuits and Science, pp. 539–544. the timing of action and perception. Current Opinion in Behavioral Pfordresher, P. Q., & Mantell, J. (2009). Singing as a form of vocal imitation: Sciences, 8,42–45. doi:10.1016/j.cobeha.2016.01.010 Mechanisms and deficits. Proceedings of the 7th triennial conference of Kuhl, P. K., Ramirez, R. R., Bosseler, A., Lin, J. L., & Imada, T. (2014). Infants’ European Society for the Cognitive Sciences of Music, 425–430. brain responses to speech suggest analysis by synthesis. Proceedings of Phillips-Silver, J., & Trainor, L. J. (2005). Feeling the beat: Movement the National Academy of Sciences, 111, 11238–11245. doi:10.1073/ influences infant rhythm perception. Science, 308, 1430. doi:10.1126/ pnas.1410963111 science.1110922 Large, E., & Jones, M. R. (1999). The dynamics of attending: How people Phillips-Silver, J., & Trainor, L. J. (2007). Hearing what the body feels: track time-varying events. Psychological Review, 106, 119–159. Auditory encoding of rhythmic movement. Cognition, 105, 533–546. doi:10.1037/0033-295X.106.1.119 doi:10.1016/j.cognition.2006.11.006 Large, E. W., Herrera, J. A., & Velasco, M. J. (2015). Neural networks for beat Pollok, B., Rothkegel, H., Schnitzler, A., Paulus, W., & Lang, N. (2008). The perception in musical rhythm. Frontiers in Systems Neuroscience, 9, 583. effect of rTMS over left and right dorsolateral premotor cortex on doi:10.3389/fnsys.2015.00159 movement timing of either hand. European Journal of Neuroscience, Large, E. W., & Snyder, J. S. (2009). Pulse and meter as neural resonance. 27, 757–764. doi:10.1111/j.1460-9568.2008.06044.x Annals of the New York Academy of Sciences, 1169,46–57. doi:10.1111/ Press, C., & Cook, R. (2015). Beyond action-specific simulation: j.1749-6632.2009.04550.x Domain-general motor contributions to perception. Trends in Lerdahl, F., & Jackendoff, R. (1983). A generative theory of tonal music. Cognitive Sciences, 19, 176–178. doi:10.1016/j.tics.2015.01.006 Cambridge: MIT Press. Prinz, W. (1997). Perception and action planning. European Journal of Liberman, A. M., & Mattingly, I. G. (1985). The motor theory of speech Cognitive Psychology, 9, 129–154. doi:10.1080/713752551 perception revised. Cognition, 21,1–36. doi:10.1016/0010-0277(85) Rauschecker, J. P. (2011). An expanded role for the dorsal auditory path- 90021-6 way in sensorimotor control and integration. Hearing Research, 271, Lombard, E. (1911). Le signe de l’élévation de la voix. Annales des Maladies 16–25. doi:10.1016/j.heares.2010.09.001 de L’Oreille et du Larynx, 37, 101–119. Repp, B. H. (2005a). Rate limits of on-beat and off-beat tapping with London, J. (2004). Hearing in time. New York, NY: Oxford University Press. simple auditory rhythms: 2. The roles of different kinds of accent. Manning, F., & Schutz, M. (2013). “Moving to the beat” improves timing Music Perception, 23, 165–187. doi:10.1525/mp.2005.23.2.165 perception. Psychonomic Bulletin & Review, 20, 1133–1139. doi:10.3758/ Repp, B. H. (2005b). Sensorimotor synchronization: A review of the tapping s13423-013-0439-7 literature. Psychonomic Bulletin & Review, 12, 969–992. doi:10.3758/ McAuley, J. D., & Jones, M. R. (2003). Modeling effects of rhythmic context BF03206433 on perceived duration: a comparison of interval and entrainment Rizzolatti,G.,&Craighero,L.(2004). The mirror-neuron system. Annual Review of approaches to short-interval timing. Journal of Experimental Neuroscience, 27,169–192. doi:10.1146/annurev.neuro.27.070203.144230 NEUROCASE 565

Ross, J. M., Warlaumont, A. S., Abney, D. H., Rigoli, L. M., & Therrien,A.S.,Lyons,J.,&Balasubramaniam,R.(2012). Sensory Balasubramaniam, R. (2016). Influence of musical groove on postural attenuation Q15 of self-produced feedback: The Lombard effect sway. Journal of Experimental Psychology: Human Perception and revisited. PLoS One, 7, e49370. doi:10.1371/journal.pone.00 Performance. Advance online publication. doi:10.1037/xhp0000198 49370 Schubotz, R. I., Friederici, A. D., & von Cramon, D. Y. (2000). Time percep- Tian, X., & Poeppel, D. (2010). Mental imagery of speech and movement tion and motor timing: A common cortical and subcortical basis implicates the dynamics of internal forward models. Frontiers in revealed by fMRI. NeuroImage, 11,1–12. doi:10.1006/nimg.1999.0514 Psychology,1–166. doi:10.3389/fpsyg.2010.00166 Shin,Y.K.,Proctor,R.W.,&Capaldi,E.J.(2010). A review of contemporary van Noorden, L., & Moelants, D. (1999). Resonance in the perception of ideomotor theory. Psychological Bulletin, 136,943–974. doi:10.1037/ musical pulse. Journal of New Music Research, 28,43–66. doi:10.1076/ a0020541 jnmr.28.1.43.3122 Skipper, J. I., Nusbaum, H. C., & Small, S. L. (2005). Listening to talking Vuust, P., Ostergaard, L., Pallesen, K. J., Bailey, C., & Roepstorff, A. (2009). faces: Motor cortical activation during speech perception. NeuroImage, Predictive coding of music: Brain responses to rhythmic incongruity. 25,76–89. doi:10.1016/j.neuroimage.2004.11.006 Cortex, 45,80–92. doi:10.1016/j.cortex.2008.05.014 Sperry, R. W. (1952). Neurology and the mind-brain problem. American Wekselblatt, J. B., & Niell, C. M. (2015). Behavioral state—Getting “In The Scientist, 40(2). Zone.”. Neuron, 87,7–9. doi:10.1016/j.neuron.2015.06.020 Stasenko, A., Bonn, C., Teghipco, A., Garcea, F. E., Sweet, C., Dombovy, M. Wilson, S. M., Molnar-Szakacs, I., & Iacoboni, M. (2008). Beyond superior … Mahon, B. Z. (2015). A causal test of the motor theory of speech temporal cortex: Intersubject correlations in narrative speech perception: A case of impaired speech production and spared speech comprehension. Cerebral Cortex, 18, 230–242. doi:10.1093/cercor/ perception. Cognitive Neuropsychology, 32,38–57. doi:10.1080/ bhm049 02643294.2015.1035702 Wilson, S. M., Saygin, A. P., Sereno, M. I., & Iacoboni, M. (2004). Listening to Stasenko, A., Garcea, F. E., & Mahon, B. Z. (2013). What happens to the speech activates motor areas involved in speech production. Nature motor theory of perception when the motor system is damaged? Neuroscience, 7, 701–702. doi:10.1038/nn1263 Language and Cognition, 5, 225–238. doi:10.1515/langcog-2013-0016 Wolpert, D. M., & Flanagan, J. R. (2009). Forward models. In T. Baynes, A. Stephan, M. A., Heckel, B., Song, S., & Cohen, L. G. (2015). Crossmodal Cleeremans, & P. Wilken (Eds.), The oxford companion to consciousness. encoding of motor sequence memories. Psychological Research, 79, Oxford: Oxford University Press. 318–326. doi:10.1007/s00426-014-0568-2 Wolpert, D. M., & Flanagan, J. R. (2015). Computations underlying sensor- Stevens, K. N., & Halle, M. (1967). Remarks on analysis by synthesis and imotor learning. Current Opinion in Neurobiology, 37,7–11. doi:10.1016/ distinctive features. In W. Wathem-Dunn (Ed.), Models for the perception j.conb.2015.12.003 of speech and visual form. Cambridge, MA: MIT Press. Wolpert, D. M., & Kawato, M. (1998). Multiple paired forward and inverse Stupacher,J.,Hove,M.J.,Novembre,G.,Schütz-Bosbach,S.,&Keller,P.E.(2013). models for motor control. Neural Networks, 11, 1317–1329. doi:10.1016/ Musical groove modulates motor cortex excitability: A TMS investigation. S0893-6080(98)00066-5 Brain and Cognition, 82 ,127–136. doi:10.1016/j.bandc.2013.03.003 Woodrow, H. (1932). The effect of rate of sequence upon the accuracy of Su, Y.-H. (2012). The influence of external and internal motor processes on human synchronization. Journal of Experimental Psychology, 15, 357–379. auditory rhythm perception. (Doctoral dissertation). Aus dem Institut für doi:10.1037/h0071256 Medizinische Psychologie der Ludwig-Maximilians-Universität, München. Yang, S. C., Wolpert, D. M., & Lengyel, M. (2016). Theoretical perspectives Su, Y.-H., & Jonikaitis, D. (2011). Hearing the speed: Visual motion biases on active sensing. Current Opinion in Behavioral Sciences, 11, 100–108. the perception of auditory tempo. Experimental Brain Research, 214, doi:10.1016/j.cobeha.2016.06.009 357–371. doi:10.1007/s00221-011-2835-4 Zarco, W., Merchant, H., Prado, L., & Mendez, J. C. (2009). Subsecond Su, Y.-H., & Pöppel, E. (2012). Body movement enhances the extraction of timing in primates: Comparison of interval production between temporal structures in auditory sequences. Psychological Research, 76, human subjects and rhesus monkeys. Journal of Neurophysiology, 102, 373–382. doi:10.1007/s00426-011-0346-3 3191–3202. doi:10.1152/jn.00066.2009 Sussman, H. M. (1989). Neural coding of relational invariance in speech: Zatorre,R.J.,Chen,J.L.,&Penhune,V.B.(2007). When the brain plays Human language analogs to the barn owl. Psychological Review, 96, music: Auditory–motor interactions in music perception and pro- 631–642. doi:10.1037/0033-295X.96.4.631 duction. Nature Reviews Neuroscience, 8, 547–558. doi:10.1038/ Thaut, M. H., McIntosh, G. C., Rice, R. R., Miller, R. A., Rathbun, J., & Brault, J. M. nrn2152 (1996). Rhythmic auditory stimulation in gait training for Parkinson’sdisease Zollinger, S. A., & Brumm, H. (2011). The Lombard effect. Current Biology, patients. Movement Disorders, 11,193–200. doi:10.1002/mds.870110213 21, R614–615. doi:10.1016/j.cub.2011.06.003